Disclaimer: This article is published in partnership with Siemens. Siemens is paying for my engagement, not for promotional purpose. Opinions are my own.
Nowadays, consumers expect the ideal product in terms of size, material, shape, color, quality and other individual needs and specific requirements. Advances in technology and digitalization have created a significant market for individualized offerings. Personalization at scale has the potential to create $1.7 trillion to $3 trillion in overall new value. Capturing this value requires companies mastering the underlying technologies and enabling consumers to be their own product designers. Some popular examples:
- Sports shoes tailored to running style and color preferences,
- Cars and vehicles configured to customer needs and driving behaviors,
- Granola specially blended to consumers’ personal tastes and dietary needs.
Evermore companies offer the option of customizing consumer and industrial goods. While personalization provides huge benefits to us consumers, it presents complex challenges to the delivering companies and their production processes. Balancing this ‘Personalization vs. Complexity’ trade-off involves several aspects, such as
- cost-benefit considerations
- technology requirements
- digitalization and automation maturity
- implementation effort
- data handling and engineering capabilities
Producing companies are therefore facing pressing questions: How can a production process evolve to meet these personalization demands? How to decide on the best automation strategy? And how will the future of production ultimately look like? It essentially comes down to the question: What to automate and to which degree? It can be all too easy to get carried away with automation for its own sake, but the result of this approach is almost always projects that cost too much, take too long to implement and fail to deliver against a company’s business objectives and operations strategy.
Personalized production – challenges and requirements
To remain competitive and keep pace with today’s market demands, manufacturing companies have to fulfill two things concurrently:
- Manufacturing customized products at the same price as mass-produced ones.
- Keeping the production process sustainable and reducing waste by producing what the market actually wants and needs (as opposed to filling warehouses, certainly not being sustainable).
The extent of twenty-first-century customization and personalization also requires production businesses accommodating a large variety of versions of any given product, which must be produced in smaller batches with very short lead times. This Low Volume – High Mix production typically means lot sizes of less than 20 pieces (occasionally, down to 1 piece only) per ordered lot, involving increased complexity and administration cost. The fundamental challenge for producing companies can therefore be formulated as: Maximizing overall productivity under
- higher personalization (i.e. increasing variety of product versions)
- smaller and variable lot sizes
- shorter product life cycles (i.e. shorter time spans available to production)
Further, a factory needs to increase its flexibility to react to frequent changes in production requirements. Bottom line: factories need to find their ‘sweet spot‘ of high productivity paired with high flexibility. “Flexibility isn’t just something inevitable factories have to cope with, but also comes along with a great bonus: it ensures to produce what really matters, i.e. providing customers with personalized products they demand by utilizing resources efficiently. That’s a clear contribution to improving Sustainability.”, argues Dr. Ariane Sutor, Head of Innovation Accelerator Zero Engineering at Siemens Digital Industries.
These impending requirements call for a paradigm shift in how factories get automated, resulting in innovation of the overall production process. The solution to meet all these challenging, yet imperative requirements ultimately lead to the vision of the Autonomous Factory.
Autonomous Factory: Personalized production at scale
Before expanding on the Autonomous Factory, let’s get one foundational question straight: what is the difference between Automation and Autonomy? Automation refers to a set of related functions performed automatically by equipment. Automation assumes that the operator performs any requirements before or after the automated sequence in order to complete the task. Autonomy, in turn, refers to a state of equipment in which it can perform the assigned operations under defined conditions without human input or guidance.
With that said, the vision of an Autonomous Factory can be delineated as follows:
- Driverless transport and smart handling systems moving products or raw materials from one point to another
- Autonomous robots performing production steps independently, without human intervention
- Flexible, modular production displacing planned-out, overarching process (as used today)
- Envisaging, simulating and implementing the optimal production path in real time
- Changing over production to an entirely new product overnight without wasting valuable time on engineering and commissioning a new production line
According to Ariane Sutor, the Autonomous Factory is built on two key features:
Optimal orchestration of autonomous machines: Autonomous machines decide independently how to accomplish a specific production task. They do not need a detailed top-down programming, but rather use a ‘world model’ or ‘Digital Twin’ to act autonomously. This hides a lot of complexity within a single machine and enables a modular approach. On the factory level, the highest overall productivity is ensured by optimizing collaboration and interaction between individual machines, leading to the most productive sequence of production tasks.
Easy Engineering: High flexibility entails frequent changes in production requirements. The resulting re-engineering effort needs to be kept as low as possible for personalized production at scale to be viable. The above-mentioned modularity provides the condition to apply a cloud-based two-part solution, pairing an ‘autonomous production engineer’ with an ‘autonomous production dispatcher’ (see visual). This combination allows production planning to be flexibly adapted to specific product requirements while minimizing time and effort needed. Siemens calls this Engineering to Zero.
It’s important to keep in mind: the more personalized and flexible companies want to serve their customers in the marketplace, the more complex their production is going to become. The main reasons for this include vast amounts of different systems interacting and communicating with each other, as well as enormous volumes of resulting data that need to be processed. Mastering this complexity entails managing a couple of key technologies, which are at the heart of the Autonomous Factory:
- Artificial Intelligence (Neuronal Networks, Knowledge Graphs)
- IIoT (Cloud and Edge Computing)
- Digital Twin (‘world model’)
As simultaneous implementation and seamless interplay of these technologies take learning curves of many years, fully autonomous factories are still a vision of the future. However, many companies are already taking progressive steps to bring this vision to life.
The Autonomous Factory is built on two key features: Optimal Orchestration and Easy Engineering. — Dr. Ariane Sutor
Outlook: How Siemens is accelerating personalized and autonomous production
A couple of factories already exhibit a relatively high level of autonomy on the shop floor (e.g. in the automotive or electronics industry), but the majority still operates on a rather limited level. Companies should start off deploying autonomous solutions in selected areas of their production that already require high flexibility. Siemens is currently following two major strategies to accelerate autonomous production and help its customers to step up their levels:
Implementation of autonomous solutions in Siemens plants. Cases in point:
Collaboration with Siemens customers to co-create innovations. Two success stories:
- Porsche: Co-creation of the new Taycan electric car’s production process, in which autonomous guided vehicles take the car bodies from one processing station to the next.
- Cloostermans: Development of an AI-based packaging solution that can respond quickly and flexibly to changing package content.
Implementing autonomous solutions is far from easy, but will become a common attribute of the ‘future of production’. Both strategies help create maximum value through tailored solutions, fostering trust in them and ensuring an optimal implementation.
Personalized products are in high demand these days. Meeting this demand is leading companies to increasingly automate their production processes and even make parts of it autonomous. However, this approach presents a trade-off: with increasing personalization comes increasing complexity. Therefore, companies need to decide on the expedient extents and levels of automation to be implemented in their factories. Two strategies that may help along the way: 1. Limited implementation in selected areas. 2. Co-creation with trusted partners.
One thing is for sure: Personalized production and Autonomous Factories will be the future – companies better set out for it now!